Supercharging Kubernetes Fury Distribution's Networking OSS Module with Cilium

SIGHUP -

It’s been two months since SIGHUP and Isovalent officially announced their partnership.
In the announcement article, we explain our common vision, plans, and goals, and we anticipate our project to supercharge and include in KFD the Cilium CNI as an additional option for our user base.

Why did we decide to add Cilium to our networking module?

Cilium is an open source, cloud-native solution for providing, securing, and observing network connectivity between workloads, fueled by the revolutionary Kernel technology eBPF that can run sandboxed programs in a privileged context such as the operating system kernel.

Thanks to these specific features, Cilium easily became one of the most relevant and reliable CNI solutions.

As we progressed in our mission to empower organizations with the best in class Cloud Native open source solutions strictly related to the CNCF official landscape, we always need to keep our open source reference architecture relevant for our customers and for everyone who decides to use it in its own production-grade environments.

Kubernetes Fury Networking

Kubernetes Fury Networking implements in-cluster networking functionality for the Kubernetes Fury Distribution (KFD) via Container Network Interface (CNI) plugins. Kubernetes has adopted the CNI specification for managing network resources on a cluster.

Historically, Kubernetes Fury Networking use CNCF-recommended Project Calico, open-source networking and network security solution for containers, virtual machines, and bare-metal workloads, to bring networking features to the Kubernetes Fury Distribution.
But recently, to keep our solution relevant, we added to the Fury Networking Module a specific package for running Cilium as CNI, and the usage of the Cilium package is specifically advised for clusters with < 200 nodes ( the use of Clalico is recommended for clusters with < 50 nodes).
The deployment of Cilium consists of a DaemonSet running on all nodes, and a operator Deployment.

Additionally, we deploy the Hubble component as an observability tool on the network connections between pods in the cluster.
Hubble is an observability tool designed for Cilium, providing real-time visibility and monitoring for network traffic in Kubernetes clusters. It collects and analyzes network flow data, offering insights into microservice communication.
With Hubble, you can visualize network activity and metrics, including network policies and service dependencies.

Conclusions

To keep our technological stack relevant and tight to the latest CNCF landscape technology, we refactored our KFD Fury Networking module, adding a Cilium package for CNI networking purposes, which we suggest using for clusters with less than 200 nodes.

The main highlight of the Cilium package are:

  • Packaged via Kustomize
  • Core Cilium + Hubble deployment available
  • Monitoring preconfigured
  • Grafana dashboards
  • All the useful metrics
  • Images hosted on the SIGHUP registry as all the other KFD modules

We've got you covered!
Check how we can help you to boost your Cilium adoption